Minimum Travel Time Tab

The minimum travel time (MTT) feature is a two-dimensional fire growth model (Finney, 2002). It calculates fire growth and behavior by searching for the set of pathways with minimum fire spread times from point, line or polygon ignition sources. In theory, the results are identical to wave-front expansion used in Farsite with the exception that all weather and fuel moisture conditions are held constant over time with MTT but are allowed to vary with time in Farsite.

Ignitions

From File

Ignitions can be shapefiles created elsewhere or within FlamMap by selecting one of the Option > Pointer Mode > Create... options.

Click the button to display three Ignition File options,

• Open Ignitions File ... - to load an existing shapefile for the MTT ignition,
• Load Current Shape - uses the current shape drawn on the display as the ignition, this option will also save the ignition as a shapefile,
• Clear Ignitions File - removes the current ignition from the run.

Random

In addition to calculating fire growth for single or multiple ignitions, MTT can be used to compute burn probabilities for a specified number of randomly located ignition points each of a constant duration. Selecting the Random radio button requires an additional input for the Number of Random Ignitions: spin box. The default number of 100 ignitions is likely too low for most landscapes. Considerations when determining the number of random ignitions to use;

• Setting the number of ignitions requires some trial and error, for best results the output map should contain only minor amounts of zero probability, ideally only the non-burnable areas.

• Since the size of each random ignition (thus it's contribution to the probability) also depends on fire duration, specified in the Maximum Simulation Time spin box, simulation times may affect results if a relativity small number of ignitions is used.

• Larger landscapes will require more ignitions than smaller ones.

• Landscapes with faster rates of spread will need fewer ignitions than "slower" landscapes of the same size.

• Since the ignitions are random, separate runs with the same inputs will give different results.

• Numbers of ignitions greater than that needed to burn the entire landscape do not significantly change the results (differences do occur with the random variability of the ignitions) but do increase processing times.

Burn probabilities provide one method of evaluating the effectiveness of fuel treatments that removes the uncertainty of ignition sources. A large sample size of ignitions (say 1000’s or 10,000’s) on the treated and pre-treatment landscape gives an indication of the overall effectiveness of the landscape pattern in retarding the growth of large fires. To do this type of analysis you will need to create different Landscape (.LCP) Files that reflect the different treatments in terms of the five fuel themes. If Random ignitions are selected only Burn Probabilities, Perimeters, FLP, and Fire Size List outputs are available.

Note: this process is very CPU intensive when simulating fire growth from thousands of random ignitions. This is best performed on multi-processor computers or when long computing times are acceptable.

From Fire List File

Burn probabilities can also be computed from a previously defined list of fires contained in a text file. This Fire List File is typically generated from a previous MTT run using random ignitions or a FSim large fire simulator run and is named the FireSizeList (because it also contain information on the sizes of the random fires).

Computing burn probabilities from a fixed list of fires eliminates the variability between runs caused by the random selection of ignitions. This can be an advantage when comparing burn probability results from runs using different input parameters. Make sure the list of fires is large enough to give good burn probabilities - typically most of the burnable landscape should end up with a burn probability greater than zero.

Click the button to display the Fire List File options,

• Open Fire List File ... - to load an existing Fire List File,
• Clear Fire List File - removes the Fire List File from the run. 

Inputs

To use the minimum travel time feature, you must supply several inputs in addition to those required for basic fire behavior computation (i.e. initial fuel moisture file, wind speed and wind direction). These additional inputs include:

Resolution of the calculations

This is independent of the resolution of the landscape data and dictates the spacing of nodes used to calculate minimum travel times and output grids. Below is shown a 30m fuel grid overlaid with a 60m node resolution. MTT performs the fire behavior calculations at the nodes.

This setting is the most significant way to control the computational intensity (i.e. run time) of your simulation. For example, doubling the resolution from 30m to 60m would decrease run time by 75% since there would be only 25% of the nodes to preform calculations on. However setting the resolution greater than the Landscape (.LCP) File cell size will cause some of the cells to be "skipped over" for the calculations. Setting the resolution smaller than the cell size provides no additional precision and only increases the computation time.

The Resolution distance units (meters, feet, or kilometers) are the Grid Distance Units of the Landscape (.LCP) File.

Maximum Simulation Time

This specifies the duration (in minutes) of the fire growth calculations for the set of constant fuel moisture and wind conditions entered on the Inputs tab. For example, if you want to simulate a fire for three days where the burning period is five hours/day (for which you specified maximum burning period conditions), enter 900 minutes. (3 days x 5 hours/day x 60 minute/hour)

A "0" (zero) simulation time behaves differently depending on the ignition type. Using a shapefile as the ignition indicates that you want to run the simulation until all burnable nodes on the landscape have "burned". On "slow" or large analysis areas or with moderate weather conditions this can cause very long runtimes.

With random ignitions or a list a zero simulation time means exactly that. An ignition shapefile is created but no spread occurs.

Interval for Minimum Travel Paths

This controls how the program selects and outputs the MTT Major Paths at a specified distance interval. Shown below are 600m (left)and 300m intervals for the major paths. This setting changes only the display of the Major Paths, it does not affect the calculation of the Influence Grid, from which the Major Paths are derived.

The Interval distance units (meters, feet, or kilometers) are the Grid Distance Units of the Landscape (.LCP) File.

Spot Probability

Launching embers from every crown fire node can be computational intensive, resulting in a long running simulation which often doesn't change results. The Spot Probability: setting controls how many of the nodes where crown fire is initiated actually launch embers which can start new fires. A setting of 1.00 will compute a spot distance and direction for each crown fire node and a setting of 0.00 will prevent any spot fires. Settings between 0.01 and 1.00 proportionally control how many crown fire nodes launch a spot. See Spotting (technical documentation) for more information.

NOTE: Users will probably want to set Spot Probability: higher for FlamMap than for Ember Spot Probability: in Farsite or NTFB fire simulators.

Spotting Delay

The Spotting Delay: setting is the time delay between when a ember lands on a burnable node and it begins to spread. This is typically set between 0-60 minutes with shorter delays associated with drier burning conditions.

Lateral & Vertical Search Depths

When searching for the fastest path MTT cannot realistically look at every possible path between nodes. To limit how far MTT searches for a faster path the Lateral & Vertical Search Depths control how long MTT will look for a faster (lower arrival time) node. Search depths are distances (number of cells) that the MTT algorithm uses to determine when to terminate searching in the vertical (among rows) and laterally (within columns) directions of cell nodes on the landscape.

They are thresholds for sequentially failing to find the soonest arrival time. For example, if the lateral search depth is set to "3", MTT will keep searching farther and farther to the right or left from the starting point, as long as it succeeds in finding a faster path, until it fails 3 times in sequence. Larger values will cause longer computation times because it will require more sequential failures before it gives up in each direction. The way the algorithm is structured, the lateral search depth is better set higher than the vertical search depth (require more columns before failure than rows).

For homogenous landscapes, larger values will increase the chance that the fastest paths are found between cells (nodes) with farther separation. In other words, MTT will be more accurate. For very heterogeneous landscapes (i.e. every cell has different spread rates or maximum spread directions) then the settings for search depths will matter much less. To increase speed for very heterogeneous landscapes, settings might be reduced to Vertical=2 and Lateral=3.

Wind direction (on the Inputs Tab)

When using the MTT feature one of the Winds radio buttons should be selected on the Inputs tab. This wind direction should reflect the results of a separate wind analysis done to find the prevailing wind speeds and directions.

Barriers

Shapefiles can be used in FlamMap runs to simulate barriers to fire spread that may not be represented in the landscape. For instance, if you know that a creek or road is a barrier to surface fire spread, but landscape data show that some or all of the cells are burnable in the area represented by the creek, you could create a barrier file. Keep in mind that fire barriers will block surface fire spread. However, torching trees could cause the fire to spot over the barrier (in reality as well as within the fire behavior model).

Barriers can be shapefiles created elsewhere or within FlamMap by selecting one of the Option > Pointer Mode > Create... options.

Click the button to display the Barrier File options,

• Open Barriers File ... - to load an existing shapefile as a barrier,
• Clear Barriers File - removes the current barrier from the simulation.

Selecting the Fill Barriers check box will fill a polygon barrier with unburnable cells so spots cannot start new fires inside the polygon. This option has no effect on a line barrier.

Outputs

The output grids are calculated at the cell size specified in the Resolution of calculations spin box described above. The MTT method produces output similar to that produced by the FARSITE fire growth models with the addition of data on the routes taken by fire spreading from one node to the other.

When burn probabilities are calculated from either Random ignitions or From Fire List File the following outputs are  available.

Four default grids are included in every FlamMap run, three of Elliptical Dimensions and one for Maximum Spread Direction. These are intermediate products of the MTT model and made available for research and creating unique output products.

Status Bar

The FlamMap "Run:" dialog box/tabs have a status bar and functional buttons at the bottom to help you keep track of where you are at with the set up process.